Duodecimal

Why'd you use rank 1? The only difference between ranks is the duration, not the chance to resist. What's the duration on Rank 1? If the duration on Rank 1 is substantially shorter than max rank, the difference between 16 seconds and 12 seconds may be 50% and 75%.

Once I get home I'll test this myself with the longest duration available (for granularity).

 

Rouncer

The one that sold me was the 2 second break.

Reason why I choose rank 1...rank 1 is 20 seconds and max rank is 50 seconds.

Chance to resist for an even con mob is 4% with no hit gear and I really didn't want to be testing this theory out for hours at a time. If the resists are based on partial time determined at time of cast then it should break at 15, 10 or 5 seconds. First break was right around the 15 mark but then I had a break that was clearly at 8 seconds left on the duration and another clearly at 2 seconds and according to the partial resist theory that should not be the case.

Also polymorph is a lot like chill or the application of a DoT. There is no way to partially resist being sheeped. You are either a sheep or you are not a sheep and that implies that it should be binary resist calculations.

Also I seem to recall back in the original alpha/beta that there was some discussion from the developers about an attempted change to the mechanic of polymorph to check every second for resistance in one of the patches and it was too prone to breakage so they went back to 2 seconds with the next patch. That was 3 years+ ago so not sure how reliable my memory is of the specifics.

If you want to test max rank please go right ahead. Although I think the level 8 alt concept would be the easiest to test since they should be able to go after a level 11 mob easily enough and look at the breakage rate.

 

Duodecimal

That's a different type of resist though. Early breaks are from Resistance, but full resists would generally be level-based down to the 1% chance to full resist. It won't help to use a level 8 alt vs. level 11 mobs because then you'd be getting the typical 17% full resist rates.

The Coilfang Guardians are notorious for breaking early in my experience. The 1% chance to resist they have left over after accounting for +hit gear has no effect on early breaks, that's just a full resist from the cast failing to land.

From my experience, there seems to be certain points at which I expect a mob to break. I've gotten in the habit of pre-casting Polymorph. This is probably just paranoia, but I've had a few irrational habits before that turned out to be based on actual game mechanics (such as jumping around like a jackass to get more range on Arcane Explosion -- I never heard that that was actually beneficial until I read about the nerf, but I always felt like I got more range when I jumped).

So, when I feel like a sheep is about to break, and it does, and if I trust my gut, then there are set breakpoints where a sheep will tend to emerge from CC. Whether it's 1/8, 1/4, 1/2 or 25/50/75 would be based not on level resists (+hit) but resistance (+pen). The problem is we have no good way of telling the base Arcane resistance of any npc except through log processing the timers, and trying to account for latency in timing.

 

◊ manly

Ok so I know this is somewhat outdated now but I was waiting for the WWS parse for this week to be uploaded before posting.

Here are the numbers from firespec for this week, if you wish to compare arcane spec to the competing specs.

Wow Web Stats - najentus - #1, 1656 dps no coe (admitedly, this was very sloppy, usually everyone gets much better numbers)
Wow Web Stats - supremus - #4, 900 dps no coe (31% partial resists)
Wow Web Stats - supremus - #1, 1100 dps with coe (last week's parse)
Wow Web Stats - teron - #6, 1823 dps no coe (this is arguably my best performance ever)
Wow Web Stats - gurtogg - '#2', 1650 dps no coe
Wow Web Stats - shahraz - #2, 1300 dps no coe

I did not post Illidan because the DPS for mages on this depends upon the roles, and anyway, I am a fire mage.

Log on with different model:
1- Create a character of the desired model. Log on/off.
2- At character selection screen, select your actual character; mouseover the new, desired model character, and hold down left click; hit enter and release left click at the same time.
bug Arcane Potency only applies to the first Arcane Missile bolt.

 

Doroteasenjk

Not a bad understanding of statistics. Just posting without caffeine and thinking I could wing the calculation. I'll try not to post so ... quickly.

I'm still certain on the heartbeat (tick) check, but I can't find the Blue quote. However, the anecdotal evidence (even my own) suggests that the resist values may be different than for conventional spells.

a) I know that at level 60, Arcane Focus had a dramatic impact on the length of time that Majordomo's adds stayed sheeped.

b) At level 70, fresh into level 70 instances, sheeps would break frequently before the 50 seconds. A fellow mage shamed me into getting a sheep watch mod, and then I shamed myself into regemming/regearing to max my hit; mobs very seldom resist or break early.

c) Mob sheeps seldom don't last the whole time.

 

Rouncer

Except that Polymorph should be a binary spell and for binary spells resists are the same whether they are based on level or based on the target's resistances.

That's pretty much what people are trying to determine, is polymorph working as a binary spell with checks on each tick or is it working as a "normal" spell capable of partial damage and the partial damage being the duration of the effect.

Since I had a polymorph break at 2 seconds to go that should pretty much invalidate the notion that the duration is predetermined at time of cast.

 

Grecasi

Don't use castsequence until 2.3.2, currently it's bugged and won't give you proper results. This is mentioned in the PTR release notes. I've had stops of 5-10 seconds spamming a castsequence macro (while I was testing optimal fire gear for me right after 2.3 on Dr. Boom), "hi dead dps".
As for supremus, use adepts elixirs, elixir of empowerment and spell penetration gear (don't gimp hit). Supremus has somewhere in the range of 250 fire resist or more.

 

pyrous

So that is infact the reason I can't keep up a scorch after 8 fireballs.

I did go to a manual cast sequence on Winterchill last night and was able to pull out a 1380 DPS. I was stationary almost the entire fight, except once when I had to move out of death and decay. I landed the #1 spot for that fight. On Kaz'rogal I dipped back down into the 700's. Im guessing thats the ignite pollution still up while Im running around.

Does anyone know what the cap limit is on spellhaste?

 

Finkum

As far as I know there is no cap on spell haste. 100% haste does not turn every spell into an instant cast; it reduces the cast time by 50%. More generally, a hasted spell's cast time will be (original cast time) / (1 + %haste) e.g. with 10% haste a fireball will be 3 / 1.1 = 2.73s. Practically speaking it is impossible to get more than 400ish haste rating (~25% haste) with the current itemisation.

 

xiaoxin21

The flamethrowers in ZA gives a 400% cast time buff which will bring your pyroblast to 1.5 sec. There don't seem to be a cap which your spell can be hasten, even if there is you cant go below the global CD, so it is sort of a cap if you want to consider it.

Next time I am in ZA, I will test this concept further by using some haste gear, ashtongue trinket of insight and see if I can get 1 sec Blizzard.

 

greyberger

Only time I'll train any ranks of pyroblast is if i know i'm going to ZA before a return to pvp spec. 1.5 second pyroblast is just too much fun to pass up.

 

Darkchani

Not quite as fun as poping skull and spamming AM tbh =)

 

galzohar

Pyroblast should maybe get some talk in the main thread. My views on pyroblast:
First of all, common misconceptions:
-Pyroblast for start of the fight due to aggro? Not really, since if you're threat limited your damage amount is limited, doesn't matter if that damage comes from pyro. Of course if you open with pyro you gave the tank 6 seconds to build aggro, but it doesn't mean using pyro was good. You were simply lowering your DPS to give the tank time to get aggro, and could do the same with fireball.
-Pyroblast as mana efficiency? While I still need to verify this with numbers, you will probably generate more mana by stacking all the time you lost pyroblasting and using it for wanding/oo5sr time in the long run. So generally a bad idea. Same goes for using "on clearcasts" (on top of what had already been said about reactive clearcast usage).

So when is pyroblast useful and why should I train it?
-If you ever spec PoM, obviously, although this probably isn't a good idea to spec, but if you do have PoM having pyro is a very good idea.
-The ZA trash.
-Any mob that is targetable but not damageable for whatever reason, will not be damageable 3 seconds from now but will be damageable 6 seconds from now. Example would be a banished mob when you have a timer and you know it's going to get taunted by a tank - timing a pyroblast to hit right as the banish breaks will both give you a good damage bonus and give the tank that taunts immidiately after it some extra threat (not over your threat, but over everyone who didn't time a pyro). Same goes for mobs that trigger immunities/reflections.

So while pyroblast is hardly ever useful, it does have its uses and is definitely worth training and keeping somewhere around your hotbars.

 

Doroteasenjk

1) Pyroblast as on opener on trash? I don't think so. Any fight that lasts longer than 15 seconds should have a full stack of Improved Scorch laid down and go into your FB/FB sequence (10 seconds if you have 2 or more fire dealers: mage and fire-destro lock). The DPS (and damage) will always be higher than starting without IS.

2) PB as mana efficiency? Don't think so either. Your mana per second suffers too much, as you say.

3) PB as an opener on basilisks while you grind out your favorite food? Absolutely.

4) PB as prerequisite to Blastwave: I think this is a good crowd control option for trash. I have saved a few healers by using Blastwave on an angry mob to slow them enough for a tank. It is better than Frost Nova in many situations because sometimes they take their ire out on the rest of the caster camp if they get rooted, but BW only slows them down and doesn't change their target.

5) PB as opener for Krosh Firehand: it gives me an opportunity for a cool countdown, as I start the pull for HKM.

 

galzohar

3. No. Pyroblast slows down your grinding. Double the cast time for way less than double damage just isn't worth it. I'd rather fireball for 6s and drink for 3s will do almost the same damage as pyroblast for 6s and fireballing for 3s and cost a lot less mana (which means you don't really need to spend 3s drinking if you want to look at it like that). Overall you will farm faster if you never pyroblast, especially if you have burning soul and/or gear that makes the mob die before/shortly after it reaches you anyway, which is doable even in full green "of fiery wrath" gear.

4. Blastwave definitely. Best spell for a fire mage for AOEing. Also works through spell reflects (as AOEs don't get reflected). Same can go for DB. Not to mention it's points you have no better use for in terms of increasing DPS assuming you'er already taking all the DPS fire talents.

5. Definitely.

 

Zaldinar

The concept for using it for threat efficiency early on is that it allows you to do slow damage that is very mana efficient. I'm pressed for time now (otherwise I'd show you some math), But Pyro does significantly more damage than fireball, for 75 more mana. Remembering that Pyro gets 1.15 frontend .2 backend from your +damage, and both ends benefit from multiplicative gains. To throw some junk numbers out there, lets assume that fireball does 100 DPS, and pyroblast does 66 DPS. Over a 6 second period you do 600 damage with fireball, and 396 damage with pyroblast, Thus you are doing less damage, and less threat, but are still doing damage vice standing around like a dink, and doing it more mana efficiently than with Fireball (Cast, stand, cast, stand, etc)

Choosing spells on clearcasting is a terrible idea, this has been shown repeatedly, unless it was procced by AM/Blizzard there is no reason to actively select those spells. That being said, unless I misunderstand the five second rule (where you get a tick after five seconds being out of combat being my understanding, if its actually five seconds out starts a two second timer, which I find unlikely since then it would be a seven second rule), you get a tick from it anyway, I'll have to confirm this later, might be worth posting in a research thread at some point. Part of the point of mana efficiency is that the DPM from pyroblast is significantly higher than Fireball, the DoT is not to be scoffed at or ignored. Math to follow this afternoon probably after work.

To truly model the game, we first must research it.
http://forums.worldofwarcraft.com/th...icId=109841969
Proven TheoryCrafting Stuff, chain casting in a PTR near you soon.

 

galzohar

When you do your math remember that while casting a pyro is a tick OO5SR, adding up those periods of 3s you lost by casting pyro and bunching them together will net a lot more OO5SR ticks. Though eventually I suppose it really depends exactly how much extra mana efficinecy pyroblast gets, which I hadn't calculated (as tbh, in practical fights you shouldn't be caring about mana efficiency anyway, and for aggro on the start of the fight, usually by the time I'm in position tank had already received a misdirect or 2 and is way ahead of my threat).

Also even if pyroblast would be worth it for mana efficinecy when threat generation is low, it lowers your DPS so much that your threat will be quite lower than the tank's (and thus you would lose DPS unless you can do higher TPS than the tank in general), so only situation I would possibly consider pyro opener for threat is if you actually could start casting before the tank engaged, yet you know exactly how much time it'll take him to engage and build threat. Overall not very practical.

 

Copernicus

It's possible to generate mana, through spirit ticks, by going Clearcasted Fireball -> Pyroblast. Since the five second rule doesn't start until a spell uses mana, that gives 3-4 seconds outside of the five-second rule. I'm pretty sure that hurts your DPS, but I have no idea what the math is because I never use Pyroblast.

 

Zaldinar

So I guess I have more time this morning than I thought I did. Doing a quick two second test of fireballing a wolf outside Ironforge then starting a pyro on a nearby boar confirms 1 tick of spirit mana regen occurs while casting the fireball. So, doing up the model for the damage of each spell (I'll use my piddly gear as a reference just to have numbers)

Stock 10/47/3 + 1 build, applicable talents are Firepower, Playing with Fire, Empowered Fireball, Pyromaniac, and Elemental Precision (We'll ignore crits at the moment). Wearing +878 fire damage, and with spirit of 207 (+64 mana per 2 seconds outside of casting)

So, stock fireball rank 13 (Thottbot World of Warcraft: Fireball) is 633 - 805 damage + 84 damage over 8 seconds for 425 base mana. Noting that the fireball DoT does benefit from multiplicative gains, but does not benefit from +damage.

( ( 633+805 ) / 2 + ( ( 3.5 / 3.5 ) + .15 ) * 878 ) * 1.1 * 1.03 = 1958.62 damage
84 * 1.1 * 1.03 = 95.17 DoT
425 * .97 * .97 = 399.88 Mana

You'll get one tick out of the fireball DoT by spamming, so:

95.17 / 4 = 23.79 damage done

Overall DPS becomes:

(1958.62 + 23.79) / 3 = 660.8

And MPS becomes:

399.88 / 3 = 133.29

With an overall DPM of:

(1958.62 + 23.79) / 399.88 = 4.96



Doing the same for Pyroblast Rank 10 (Thottbot World of Warcraft: Pyroblast) 939 - 1191 damage + 356 damage over 12 seconds for 500 base mana. 1.15 frontend .2 backend coefficients, and the DoT benefits from multiplicative gains.

( ( 939 + 1191 ) / 2 + (1.15 * 878) ) * 1.1 * 1.03 = 2350.64 damage
( 356 + ( .2 * 878 ) ) * 1.1 * 1.03 = 602.3 DoT
(500 * .97 * .97) - 64 = 406.45 Mana
(2350.64 + 602.3) / 6 = 492.16 DPS
406.45 / 6 = 67.74 MPS
(2350.64 + 602.3) / 406.45 = 7.27 DPM


So, looking at this, you have a gain of 3.31 DPM (significant) for a loss of 168.64 DPS (also significant). However, comma, when using as an opening shot you can start three seconds earlier, and blow fireballs DPS and DPM out of the water, which negates the concept of using it for threat efficiency. But strike for strike it is *incredibly* more effective (46.57% more effective in my gear and scenario) than fireball for damage per mana.

I think the differing assumption that we're making is that the general conception is that DPS is always the most important thing to keep in mind in all situations, which is not entirely accurate. I know in a perfect raid environment it should be, but as I have encountered in the past repeatedly the perfect raid encounters don't always exist.

A great example being a few Kara (yeah I know, get out of Kara and do BT before you talk about raiding, yadda yadda) encounters I've been in, where one of our DPSers kept getting killed by environmental effects, the loss of a chief DPS actor in an encounter forced me to change pace from speed to longevity, which I accomplished by using Pyro in the rotation, put it out, let the DoT run its full duration, eat up the DPM benefits. In a scenario where your mana bar won't be exausted by the encounter and you wont outrun the tanks threat output then yes, Pyro doesn't have a place in your rotation other than if you want a safety bubble when starting early by slowing your DPS, which you could accomplish either with pyro, or a few extra scorches, or wanding as has been suggested.

To truly model the game, we first must research it.
http://forums.worldofwarcraft.com/th...icId=109841969
Proven TheoryCrafting Stuff, chain casting in a PTR near you soon.

 

◊ Vontre

Just use scorch if you need mana efficiency. I hear it also stacks this sweet debuff!

www.magegraf.com

Raiding is full of challenge. Sometimes there is fire. You have to not be in the fire.

"We agree with Communism." - Greg Street 2009

 

Sarastro

If the numbers Dr. Damage give me are correct, Scorch is actually less mana efficient than Fireball. Scorch uses less MPS though. I guess that would have to do with Fireball getting 115% of spell damage (w/Empowered Fireball), while scorch only gets 43%, which is obviously not made up by the additional Scorch crit from "Incineration".

Of course you were just pointing to the fact that it makes more sense to stack the scorch debuff while the tank is building threat, instead of casting pyroblast.

 

• Arawethion

I've always found MPS as a more coherent way to compare things. The problem with trying to evaluate DPM directly is that there are a lot of constant offsets to you mana consumption, based on MP5, VT, etc. Generally, when you're short on mana, what you need is a low MPS spell like Scorch, regardless of the apparent damaga/mana from the tooltip.

Answers to Moonkin questions:
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

◊ Vontre

Huh, so it is. Guess I hadn't checked that since WoW1.0... scorch is very, very slightly less mana efficient. Interesting.

Edit: If you don't count VT, that is, so hrrm.

www.magegraf.com

Raiding is full of challenge. Sometimes there is fire. You have to not be in the fire.

"We agree with Communism." - Greg Street 2009

 

Sarastro

Before 2.3, Fireball (in a 10/48/3 build) only had 105% damage coefficient, the coefficient un-nerf probably turned it around.

 

Muphrid

I personally prefer to define damage per mana = damage per second / net mana per second. Net mana per second incorporates regeneration effects. You can, of course, define it as damage per cast / net mana per cast, but since most mana regeneration means are rates anyway, this seems a bit backward.

Defining "real DPM" in this way also has the advantage of making your damage per mana pool clear: it is simply (size of mana pool)*("real DPM"). Of course, damage per mana pool in and of itself may not be useful, as even after you go completely OOM, you regen and eventually rejoin the fight.

Now, I don't know how Dr. Damage chooses to calculate DPM, so this may be moot.